deepskyfrontier: January 2010

Sunday, January 24, 2010

I Am Not A Hoax

Someone sent me this link calling into question whether the dvorak ergonomic keyboard arrangement is what people (like me) claim it to be.

Here's a quote from the last page of the article:

The QWERTY keyboard cannot be said to constitute evidence of any systematic tendency for markets to err. Very simply, no competing keyboard has offered enough advantage to warrant a change. The story of Dvorak's superiority is a myth or, perhaps more properly, a hoax.

Here's my story. I learned to type in my first semester as a high school freshman. It was a once-a-week class. We used WordPerfect 5.0, which runs in DOS. Unlike Barbara Blackburn, one of the fastest typists to ever live, I did not fail my high school typing class. At the end of the semester, my teacher told me it was okay to stop trying after I hit 60wpm. And that made mad. I retook the test and turned in a score of 75wpm.

I did a lot of writing when I was in high school. Millions of words, tens of thousands of pages. All of it on qwerty.

Somewhere along the line I learned of the existence of dvorak. And when I did, I felt cheated. You see, qwerty is all fine and good once you learn to touch type, but it's inherently inelegant. So, at some point in my first semester of college, I started practicing on dvorak. Despite the interference posed by knowing qwerty, I was able to reach 20wpm in a couple hours. At the beginning of my second semester of college I realized that I was going to be doing a lot of writing in my life. I decided to switch to dvorak. So I wrote a clumsy macro keyboard for WordPerfect 5.0 and I quit qwerty cold turkey. It was frustrating at first. My mind would wander just a little and I'd find myself making qwerty-specific mistakes. But I'd correct them immediately. I didn't learn dvorak in a class. I learned it a couple hours and then practiced it in real life for months before I had matched my speed to that of qwerty.

Now there are times when I use qwerty extensively, even to this day. In fact, if I'm writing an email- which is rarely terribly long- qwerty doesn't irk me. And if I'm using a program with shortcuts, qwerty makes more sense. In fact, I don't even bother to rearrange the keys on my keyboard.

However, when I'm writing, sometimes for eight or ten hours at a stretch, dvorak is indispensable. Not because I can type much faster with dvorak (I can hit over 100wpm when I'm racing- something I've never come close to on qwerty), but because dvorak is effortless to use.

Let me give you one tiny example. The word "THE."

On qwerty, "THE" is typed by leaping diagonally with the left pointer finger, side-stepping slightly with the right pointer finger, and then reaching up one row with the left middle finger. Alternating motion. No great distances to travel.

On dvorak, "THE" is typed by pressing the "TH" combination- right middle then pointer in quick succession- no traveling- and then pressing straight down with the middle (strongest) finger of the left hand. Is it faster than qwerty? No and yes. No, it's not faster. At normal typing speed, it's just easier, relatively effortless. Does that make it more satisfying? Sometimes. It actually depends on my mood. Sometimes I like all the work my fingers are doing. I like that they're jumping all over the place. But after about half an hour, that gets old. And after two hours, it gets very old. So, eventually, the answer is absolutely "Yes." Dvorak is faster. But it's also faster if what you're going for is raw speed. If I'm copying something in front of me, or if I were taking dictation, or doing Close Captioning (CCers use dvorak if that tells you anything), then dvorak would allow me to type much faster than I'd normally need to if I were just trying to keep up with my own thoughts.

And there are kinds of writing for which dvorak allows my speed of thought to increase.

You may be thinking, "That's great if all you have to type is the word 'THE'." I mention THE only because it symbolizes the entire experience of typing on dvorak very succinctly. On dvorak, my hands don't move as much. I just counted how many times my fingers left the home row in the course of a single sentence. On dvorak: 23% of the time. That's right. 77% of the keys of I typed didn't involve leaving the home row. That's because, on dvorak, the home row represents the most common letters. How common? They make up about three quarters of the words that make up the English language. Especially the common words. How does dvorak achieve this feat? Here's a hint. On qwerty, only one vowel, "A," is found on the home row ("E" is actually the most common vowel in English). And one of the keys on the home row isn't even a letter. It's the semicolon. And yet, every English word contains vowels. And the dvorak keyboard has all of them, except "Y," on the home row.

Same sentence, on qwerty: 63% of the time was spent off of the home row. And I never used the semicolon once. In fact, many intelligent people may legitimately go their entire lives without even feeling a need to use a semicolon. As punctuation, it's an option, not a necessity.

Does that make dvorak faster? Yes. It does. Have there been inconclusive studies, comparing average typists to people using qwerty at speeds below the threshhold above which dvorak becomes indispensible? Yes. Do these studies measure the typing samples in minutes rather than hours? Yes.

Are there people who feel that, because they type "fast enough" on qwerty, that dvorak is a needless waste of effort? Yes. Yes, of course there are. People who aren't upset at their high school typing teacher for making them learn on qwerty because, after all, it wasn't that hard to learn. Sure.

Here's the problem. I know people who don't know how to type. People who took high school typing- for one day a week x one semester- and who don't know how to touch type. They look at the keys as they type. They do an advanced version of hunt and peck. And they get by.

Recently, I heard one of these people in an interview. He was asked why he doesn't try his hand at writing a novel. His answer? "I type too slow." A lot of people have the same story. Learning to touch type is an essential skill in our society, and yet many people don't learn it.

Remember when I said that it took me only a couple hours to learn to type on dvorak- to 20wpm- and that the rest was practice? That's the truth. Beginner typists make faster progress- and it even feels faster because they're typing actual words right from the beginning. Once you break through into the realm of touch typing, the practice part comes naturally. Qwerty takes a long time to master. It takes days of primary learning just to "get" where the keys are. Break through is not inevitable. Why? Because the arrangement doesn't make sense on any level.

Well, it makes sense on one level. Try typing the word "typewriter" using the qwerty keyboard. Notice anything strange? It's not an accident.

People who write for a living, who write articles quoting research calling dvorak a myth, are using the qwerty keyboard. They're touch typists. And they don't have a clue what they're talking about.

Why hasn't dvorak been adopted? It's simple. The people that are in a position to make such a decision already do "just fine" using qwerty. They don't see it, and they can't be shown. Like the Matrix, sadly, the only way to know what dvorak is involves swallowing the red pill and learning dvorak for yourself.

There is no faith involved when I say that dvorak is superior. I am speaking from ideal experience. I can still use both qwerty and dvorak with equal facility. I prefer dvorak because it is superior for reasons directly attributable to its intentionally ergonomic design. Speaking of that article: it is frankly unbelievable that someone could seriously suggest that dvorak's superiority is a "hoax" based on the suggestion that specific missing evidence constitutes negative evidence and the fantasy that markets don't make such big mistakes. Along with most of America, I can personally testify that they do. I don't expect people who are already proficient at qwerty will be able to make the switch- and that's most people. High school typing students don't have the right to make decisions. They do what they're told, and if they don't they don't learn to type.

However, I did it myself, and it wasn't hard. It wasn't a mistake. I don't regret it. And I know what I'm talking about. Anyone with basic math skills can see that qwerty was not designed for efficiency and that dvorak was. And design matters.

Dvorak is so effortless in long writing sessions that I can literally write up-to and past the edge of sleep. I can write in a hypnagogic state and, while it doesn't last for more than a paragraph, I can actually read what I wrote the next day and be able to track the changing of consciousness: from things I can remember to things I have no recollection of. I can not do that using qwerty.

Tuesday, January 19, 2010

The Cost of Power

In 2008 the world used about 15.04 terawatts of power, on average. That means that, at any given point in time, 15.04 terawatts of power generation was needed. This includes all the electricity, all the jet fuel in the B-52s, all the cars and trucks and trains and ships, all the firewood, all the geothermal, all the energy we require to operate planet earth in a style to which we are accustomed.

For the sake of this thought experiment, let's pretend that it's all electricity. How many kWh (1000 watts for 1 hour) does that represent?

Let's do the math:

There are about 8,766 hours in a year.

15,040,000,000,000 watts x 8766 hrs = 131,840,640,000,000,000 watts.

131,840,640,000,000,000 watts / 1000 = 131,840,640,000,000 kWh.

That's about 132 trillion kWh.

There are 6.796 billion people on earth. That means that, on average, the energy needs, per person, are about 19,400 kWh / year, or 2,210 watts per hour. That's per person, not per household. For no good reason, let's say that the average cost per kWh is $0.15. That comes to a yearly energy bill of about $2,910 per person. That includes the cost of running the tractors that plant and harvest all your food, the energy cost of manufacturing the goods you buy, the trucks that deliver them, street lights, the mall's utility bill, cooking, heat, logs for the fire, electricity, jet fuel for visiting grandma, etc. Doesn't sound so high does it? But it's an arbitrary estimate. Again, the actual number is invisibly complex. Anyone who knows the answer, please educate me.

Q: Where does this power come from?

A: 37% oil, 25% coal, 23% natural gas, 6% nuclear, 4% biomass, 3% hydroelectric, 0.5% passive solar, 0.3% wind power, 0.2% geothermal, 0.2% biofuels, 0.04% solar electric.

89.2% of that comes from fossil fuels or biomass (think firewood, ethanol, biodiesel).

The rest comes from a mix of mostly nuclear and partly renewable sources.

In the U.S. the majority of our power production comes from coal. According to this creative little breakdown (which fails to take into account the cost of environmental damage done by coal mining, CO2 production and simply makes up numbers on the nuclear side), coal production is insignificantly cheaper than nuclear. Does nuclear power produce pollution that should be included in the cost? No. It produces waste that needs to be sequestered away to avoid becoming pollution. According to that web page, 17% of the cost of nuclear energy is the cost of storage. Other than that, nuclear is clean.

Let's play a little game. Let's say that all power plants cost the same amount to build. That's obviously not true. A coal plant is much cheaper to build than nuke plant. An oil-fired plant is probably cheaper still. But let's pretend that the environmental cost of coal is measured in money and that the money has to be paid when the plant is built. There, fixed.

Let's also pretend that the only kind of power we need to concern ourselves with is electricity. Is there anything that electricity can't do that other forms of power can? Yes. Electric trucks, planes, ships- no go. Hydrogen, generated from electricity, that's another thing. But not another enough. We'll assume that at least a third of the cost of energy will be tied to the price of oil until oil is no longer for sale.

What is that price?

One barrel of oil is equivalent to 6.1 gigajoules of heat energy, which is equivalent to 1,700 kWh of heat energy. About 60% of that can be converted into electricity. That leaves about 1,000 kWh per barrel of oil.

Q: To serve the entire world's energy needs by burning oil, how much oil would you need to burn and if you were to buy it all at today's price of oil, what would it cost?

A: 132 billion barrels of oil. At $79 / barrel (1/19/10), that could cost a mere $10.4 trillion. Just for the fuel. You see, if you could get your power generation for free and just pay for the fuel, electricity would cost you about $0.08 / kWh. Oil ain't cheap.

Q: Same question, now with natural gas.

A: 6000 cubic feet of natural gas is equivalent to one barrel of oil. Natural gas costs around $5.36 / 1000 cubic feet. That means that the natural gas equivalent of oil is about $32. That's $4.21 trillion for the whole world, or $0.032 / kWh at raw generation costs.

Q: Now with coal.

A: I've got bad news. Coal plants have an actual thermal efficiency of about 30%. At that rate, they produce approximately 2 kWh per kilogram of coal. That means that one barrel of oil is equivalent to about 500 kg of coal. In 2008 a metric ton of coal in the U.S. cost about $47. In much of the world it cost 2-3 times that. *Rolls Eyes* Soooooo, about $24 per barrel of oil equivalent. That means that the total raw cost would be $3.168 trillion for the world or about $0.024 for power generation.

Q: Finally, nuclear.

A: Accord to a recent article in Wired, traditional nuclear power costs about $55 million per year to run a 1 billion watts reactor. That's a barrel of oil equivalent of $6.30, which means that the whole world's power requirements would be met for $831.6 billion. And the raw kWh cost? $0.0063.

Now, as referenced above, defenders of energy-from-coal would have you believe that the actual cost is a lot closer to that of coal. Even a little higher. They're probably right. But remember, we're assuming equal cost for building the generators. Nukes are bigger and there's an economy of scale that works in their favor. But big works against local, and local is an important part of efficient, which is an important part of cheap.

Q: Same question- only now we're talking about renewable energy- solar, wind, geothermal, hydro.

A: Remember when I said we won't count the cost of building the plant? Well, if we apply that rule here, the answer comes to zero. Obviously, plants cost money.

Now for the fun part.

In the last question, I said "traditional" nuclear power. That means enriched Uranium in carefully managed reactors with massive cooling towers and- despite being massively complex-actually trade away a huge amount of efficiency so they can be made simpler. A nuke plant is the size of a fair sized town and employs as many people (per shift). It produces weapons grade plutonium as a by-product which is great for winning Cold Wars (or destroying the planet, take your pick- or, for that matter, building nuclear propulsion starships).

Liquid thorium reactors were proposed way back in the 50s, developed through the 60s and early 70s, and thoroughly forgotten for a generation. Thorium is a superior fuel to uranium. Less radioactive, but easier to sustain. And it dissolves in flouride salts. It's possible to build a passively regulated reactor in which, if the reaction gets too hot, the molten salt expands and automatically slows the reaction. Such reactors could be built very small- as little as 3000 square feet x several stories tall. They could be used to power ships. Hell, they could be used to power large aircraft. Their nuclear by-products are relatively benign compared to uranium- becoming safe in a couple centuries instead of tens of thousands of years. And they don't help you build bombs. Which is why they were never developed. But it's also why they've become very popular with growing nations like India and China, or hyper-visionary development zones like Dubai. It's why Harry Reid and Orrin Hatch- who have never liked the idea of storing nuke waste in their back yards, have championed their development. And thorium is abundant. And cheap. There are downsides. For instance, getting the reaction started is the hard part. But once started, it can be kept going for the life of the plant. According to stats quoted in Wired, a one gigawatt reactor would take $10,000 in raw thorium fuel per year.

Q: How much would it cost to provide the entire world's energy needs using thorium reactors?

A: Try this on. A barrel of oil equivalent of less than 0.12 cents. A world cost equivalent of $151 million. A per kWh cost of less than $0.0000015.

Remember back when I mentioned the worldwide per capita cost of energy? What was it $2910? Well, if there were some way to produce all our energy needs using thorium reactors, the raw energy cost would be about $0.03. That's not per kWh. That's per year.

Q: What would it mean if all our energy needs were met for less than $3 / lifetime?

A: Well, we'd change the way we use power. We'd build maglev superhighspeed trains to replace continental air travel. Transportation costs would drop- and with it the cost of goods. Tourism would benefit. Cities would get brighter. Bright enough to grow food indoors, underground, or out in the open in the middle of the night. Astronomers would hate it. We might consider using mass drivers in place of chemical rockets to launch payload into space. All our heat would be produced electrically, virtually eliminating pollution. We'd find it more attractive to live in Alaska, northern Canada, Siberia. Cheap power would make it possible for more people to have access to education, health care, and entertainment. Energy intensive desalination would be commonplace. Your TV would probably get a lot bigger, a lot brighter (though there still wouldn't be anything on).

It would be a different world, but not so different. People would still be poor. People would still waste resources. People would still find ways to damage the environment. But in many ways it would be the world predicted back in the 50s, before the Cold War drained all the fun out of the future.

Friday, January 15, 2010

Replacement Earths for $1

How big would a city have to be if it contained the entire population of the world and was comprised entirely of high rise apartments?

Let's lay out our playing pieces first.

1. The world's population is about 6.796 billion.

2. The area of all the world's land is about 57.5 million square miles.

3. The worldwide population density is about 118 people / square mile.

4. The population density of New York City is about 27,440 people / square mile.

5. The population density of Manhattan is much higher: 71,201 people / square mile.

Let's ask ourselves some questions:

Q: Name a country with a population density similar to that of the entire world.

A: Afghanistan is a good match: 118.6 people / square mile.

Q: Take all the people in New York City (8.363M) and spread them out at the same average density as the entire planet (118 ppl/sq.mi). How much space would they take up?

A: About 71,000 sq.mi. - an area about the size of North Dakota, Missouri, or the country of Cambodia, Syria, or Uruguay.

Q: Same question, only now you're taking just the people in Manhattan (1.635M) and spreading them out at the earth's average population density.

A: The population of Manhattan would take up about 13,900 sq.mi.- about the size of Taiwan or Moldova. No states are in the right range of size. Twice the size of Hawaii? Meh.

Q: What if we took those 1.635M people and spread them all over the planet? What would the average population density of Earth be?

A: 0.0284 ppl/sq.mi. Greenland, by comparison, has a population density of 0.067 ppl/sq.mi. Each person on earth would have 35 square miles all to themselves- approximately the size of the island of Anguilla or, ironically, an area about the size of Manhattan (33.77 sq.mi.) all to themselves.

Q: If you were to take the entire population of the world and pack it into a single large city with a population density similar to that of New York City (the Five Burroughs), how big would it be?

A: About 248,000 sq.mi. - about the size of Texas. And no country is the size of Texas.

Q: Same question, only now we're using the population density of Manhattan as a model. How big would our "World-Sized Manhattan" have to be?

A: 95,700 sq.mi. A little bigger than Indiana or the United Kingdom.

Q: Imagine that the entire planet is covered by an endless Manhattan-like cityscape. How many people would live in it?

A: Just over 4 trillion people.

Let's take those 4 trillion people and give them the following: 1/4th acre of farmland (in a vertical farm)- about 10k square feet. 1000 square feet of personal living space. A 2000 square foot share of open space. A 1000 square foot share of shared infrastructure space. 1000 square feet of structural space. Total area: 15,000 square feet. Let's assume a ten foot ceiling for all spaces. Total volume: 150,000 cubic feet. Of course, give that person a super-advanced MMO account and they could feel like they had a lot more space.

Q: How big would a spherical spaceship need to be to house the entire population of Planet "Super Manhattan" (4 trillion)?

A: 1046,448 feet or almost exactly 198 miles in diameter. That's a very small moon or average-sized asteroid. For comparison, Saturn's moon Mimas is about 25% larger. Ceres, the largest asteroid (now dwarf planet), is about 260% larger.

Q: What if you were to just put the present population of the planet (6.976B) in such a spherical space spaceship, only this time let's give them four times more space: 600,000 cubic feet per person.

A: The sphere would be almost exactly 200,000 feet in diameter or just under 38 miles in diameter.

Q: Don't be so generous. 150,000 cubic feet is excessive. Using aeroponics, virtual reality, and creative design, you could make a person quite comfortable in less than 50,000 cubic feet. How big would your sphere be then?

A: About 86,580 feet or 16.4 miles in diameter.

Yeah. That's a sphere, less than 17 miles in diameter, that comfortably houses the entire present population of the planet.

Q: What would it cost to build such a sphere?

A: About $1 in the year 2100. Why? Because if you can build in space and can use material from asteroids and get your energy from the sun, the only cost is labor. But if you can build a robot that can build another robot that can also build another robot- using free material and free energy- the only cost is the original robot, the original program, the cost of delivering it to a place with abundant free resources, and *time.* How long? Well, if you started at the center and built outward at a rate of one inch an hour, it would take you 115 years. So you'd have to build it faster than that. If you could build it at an average rate of an inch a minute, it would take just under two years. How much would your first robot cost? It might cost only 50 cents and weigh only a few grams- much of which would be a solar array that could double as a solar sail. If you bought it in space, you're already most of the way there. What's the other fifty cents for? I don't know. Maybe you want to buy two just in case one gets lost.

Q: What would it cost to build a bigger one?

A: Same amount. It would just take longer.

Tuesday, January 12, 2010

Crowd-Sourcing Infrastructure Inspection

I read somewhere that our aging bridges and dams are being inspected at a tiny fraction of the rate necessary to keep ahead of disaster.

So, turn basic inspection over to amateurs. Advertise for volunteers. Gather them from the local community. Don't pay them. Have them periodically take pictures of anything that looks off. Have them send these pictures to actual inspectors. Let the actual inspectors respond while things are still fixable.

It would be better than nothing. Nothing is almost what we've got now.

Wednesday, January 6, 2010

Design for a Versatile Conference Table

(Note that the lighter green rectangle under the table is there only to help visualize the position of the four anchor points).

This idea would work best with a rather large table. The concept could work with smaller tables too, but I conceived it as a way of making a large conference table more versatile.

The table surface is suspended, rather than held up by legs. It's suspended from points away from the edge of the table.

Of course, if all you did was suspend it, it would swing around like a oversized swingset. That's why it's anchored to the floor.

There are four anchor points on the floor that are set back from the edge of the table to avoid interfering with the legspace. Each anchor point has two cables attached. One that crosses to the far side of the table, and one that crosses to the far end. A total of eight cables secure the table.

By detaching all eight anchor points and raising the table ona pulley system, you could store the table on the ceiling.

To secure the anchor points, you simply lower the table below the designed height and lock it in place temporarily. You then attach the four pairs of leg anchors which, at this point, are still loose. You then raise the table, pulling the leg lines taut in the process and then lock the height.

And by installing anchor points in the floor that are positioned closer together, you could anchor the table at coffee table height. By putting them farther apart, you'd have a standing-height work table. You'd simply lock the top cables at a different height. By putting the upper suspension cables on a pair of rails and putting anchor points at different positions on the floor you could move the table from one end of a room to the other.

You could anchor the cables more directly- more or less straight down- but the length of the leg cables wouldn't allow for multiple height configurations, would interfere more with legspace, and would require higher tension (and therefore, stronger cables) to dampen lateral sway.

Tuesday, January 5, 2010

Proposal for an American Super Tower

Imagine building a city where previously there was none. Imagine building much of it as a single structure. That's what's being undertaken in China, Dubai, Saudi Arabia, and most recently, Russia. Undertaken doesn't necessarily mean "assured." Real estate can be a good investment though, and the people that tend to live in 1/2 mile high towers do tend to have some cash to spend so such plans aren't necessarily as crazy as they look and sound.

Now that the bar has been set, what would it take for the United States to play this game?

The Burj Dubai will be a science-fiction-flavored 828 meters tall when completed. Let's ask a very serious question. What would it take to build something twice that tall- a mile in height? 1610 meters. Or, while we're up there, what would it take to build something 1776 meters tall? That almost exactly 110% of one mile.

First, let's divest ourselves of the fantasy of doing it the "old fashioned" way. There are no traditional building methods that would allow such an undertaking. The methods used in the Burj Dubai are already maxed out. Otherwise, they would have gone higher.

The answer is to depart from the requirement that the building be "free standing." Instead, you support it in tension, like the third tallest structure in the world, this 2000ft+ tall radio mast.

A radio mast like this is actually an invisible pyramid of high tension lines which extend outward as odd-numbered spokes from the single central compression member of the tower itself. The need for stiffness is all but eliminated using this method, which saves an incredible amount of weight that would otherwise be required. The central mast needs only be stiff enough to resist the forces exerted by the wind operating on the spans between where the tension lines are connected. And it need only be strong enough to support its own weight and the weight of the cables pulling down on it. Such radio masts have almost no payload- just a few hundred pounds of radio antennas.

But the idea is scalable. Take a look at one of the anchors for the above-mentioned radio mast. Notice that three cables come from a single point, attaching to three elevations of the mast itself. At first glance you might think, "that's big." But really, it isn't. It's actually "just big enough." If it needed to be bigger, it would be bigger. There are no physical laws that say that you can't build a bigger anchor.

Here's what I have in mind. First, start with a 450 meter free-standing monolithic base. Most of the salable real estate will be found inside. From there, continue upward with a tapering tower that is with another 850 meters of usable and semi-usable payload- airship apartments that one would be serviced by elevators operating at near freefall speeds. The final 476 meters would be built to support a single observation deck.

The wind would exert hurricane force on the upper reaches. By placing numerous wind turbines on the upper structure, some of that lateral force could be transformed. The added weight of the turbines would cancel most of the benefit of doing this, however, what would be gained in electricity would probably power much of the structure's needs. Individual turbines might be pinwheel sized. Too large and their blade lengths would extend beyond the width of the tower itself which would actually increase the wind load.

Power could also be generated by harnessing the variable tension moments in the high tension lines. This would involve building linear hydraulic pistons into the anchors. The pistons would lift a multi-ton weight. By diverting some of the pressure thus stored, a turbine could be driven, resulting in additional energy gains. The same system could be used to stiffen or loosen one side of the tower's supports, allowing the structure to lean into the wind and thereby remain upright.

The uppermost diameter of such a structure might be only 5 meters. It would be almost entirely made of steel.

Changing the method of construction multiple times from foundation to point would be analogous to a multi-stage rocket. What works close to the ground is different from what works higher up. People buying real estate would be buying the prestige of living in the tallest building in the world, not living in the highest part of it- a situation that would be far from desirable anyway. Above a certain significant height, only the most courageous would feel comfortable anyway. And the amount of space available would diminish such that the daring billionaire might find herself living in something like a lighthouse's arrangement of stacked rooms.

I should say something about it's sensitivity to attack and how it would need to be built to accommodate certain types of failures that would be unlikely except as acts of sabotage.

The first 450 meters would be immune to all but bomb attack. Cutting all the tension lines would not affect it. The next 850 meters would be susceptible only to cutting multiple tension lines and would feature two modes of emergency escape. One mode would involve going down the elevators. Another would involve riding down the tension lines in small enclosed gondolas equipped with speed-activated brakes. The uppermost observation deck would be similarly equipped, only the gondolas would (unfortunately) be even lighter. The structure would need to be built to withstand 200mph winds and the loss of a random assortment of cables. Each anchor would be guarded. Flying an airplane into the tension lines would usually result in simply cutting the aircraft into kites. The resulting shockwave would be absorbed using the active hydraulics mentioned above.

What would it look like?

Imagine something like the Eiffel Tower, only- instead of four sides- it would have an odd-number of sides. The number of sides would gradually decrease as you got higher and higher as features in the outer facade merge. It might have a spiral appearance, much like the Crystal Tower concept. I have a quite different design paradigm in mind that might be applied- I'll elaborate in a future post. In any event, it would have perfectly smooth transitions. You could not tell by looking at it where one construction paradigm begins and another ends. And it would be lit up. The tension lines might be lit with millions of lights. Because of the foreshortening of perspective at such great distances, the top would taper such that, when one is standing withing the tension umbrella, it would appear to extend upward infinitely. The tower's actual vanishing point would be scant arcminutes from the top of the tower.

Where would such a structure be built? California is likely out of the question because of the earthquake risk. In fact, the entire West Coast might be out. Shame that. Not that CA laws would ever allow it. It should also be build away from the most common paths of hurricanes, so the Gulf Coast is also mostly out of the question. Although, it is already being built strong enough to withstand hurricane winds. What's an actual hurricane?

Building too far north involves a reduced building season. Too close to NY City would involve air traffic interference. In terms of real estate value, perhaps the best placee to build would be within sight of Manhattan. Too close to Manhattan and it would do injury to the skyline aesthetic. But looking down on Manhattan would be worth paying for. How much? Between $30k and $150k a square foot.

The location question is something that would take a lot more thought.

What would it cost to build such a structure?

Well, the design and engineering costs would come to between ten and twenty million dollars. To purchase and prepare the building site would require tens or hundreds of millions of dollars. It would incorporate around a million tons of steel (more likely more than less). That's some $1B in raw material costs. Raw construction would cost around $8B. Finishing it would cost at least another $5B. Around $16B in a perfect world. Actual cost is likely to be about twice that. To be profitable, it would need to attract businesses and residents from an area larger than its viewing radius. It would need to compete with Dubai for billionaires.